Chapter One: Introduction

Reported attitudes to the use of animals in science: Artefacts of surveys?

Julia Veitch

A sub-thesis submitted in partial fulfilment of the requirements for the award of Master of Science (Scientific Communication)

January 2002

National Centre for the Public Awareness of Science Faculty of Science Australian National University

Declaration

I certify that this thesis does not incorporate without acknowledgement any material previously submitted for a degree or diploma at any university; and that to the best of my knowledge and belief it does not contain any material previously published or written by another person except when due reference is made in the text.

Julia M. Veitch

ii

Acknowledgements

I am indebted to Dr Sue Stocklmayer, my supervisor at the Centre for the Public Awareness of Science, for her close reading of various drafts and her invaluable guidance. I would like to thank those social psychologists, both attitude researchers and discourse analysts, who corresponded with me, sent material and gave permission to include unpublished survey instruments in this thesis. I am grateful to my husband, Sagar, for the constant assistance and encouragement he has given to me for this project.

iii

Abstract The animal rights movement has brought to the attention of the public their concerns about the way animals are used in science and whether they should be used at all. The research of scientists and psychologists who use animals has been directly affected by the mobilisation of public opinion and by animal rights actions. Much research has been generated on the subject of public opinions about the use of animals in science, seeming mostly to show that a large percentage of people believe animals should not be used for scientific experiments. Material including books, research papers and survey instruments designed by social scientists on attitudes to animals generally and in scientific research were analysed discursively to see whether attitudes as measured were to some extent an artefact of the surveys themselves. The results of the analysis showed that both construction of items in the surveys and secondary sources generally show a negative orientation toward science in three major ways. First, experimental animals are described in terms that promote readers’ identification with the animals as being like themselves and therefore deserving rights, including the right not to be used specifically in the way described by the material. Second, misinformation about actual animal use in science is structured, albeit unintentionally, into survey items. Third, science is described in terms that are difficult to identify with, being either abstract or portrayed as irrelevant to every day life. The effects of the material in this study are often to misinform readers about the actual use of animals in science and the broad ethical purpose and value of the science that uses animals. The material also omits the ethical consideration of why animals are used in experiments rather than humans. Future effort needs to be put into gathering data on what people think about animals in research in a way that is both informed about the wider ethical considerations of the use of animals in science and that is sensitive to how variable and context-dependent people’s responses are.

iv

Contents Declaration

ii

Acknowledgments iii

Abstract

iv

Chapter One: Introduction 1.1

Background to the study

1

1.2

Statement of the problem

2

1.3

Purpose of the study and research questions

2

1.4

Significance of the study

3

1.5

Limitations of the study

4

1.5.1 1.5.2 1.6

Sample Technique of analysis

Overview of the study

4 4 5

Chapter Two: Review of related literature 2.1

Introduction

6

2.2

How attitudes to animals in research came to be a subject of investigation by social scientists

6

What the surveys have sought to find out

6

2.3

2.3.1 2.3.2 2.3.3 2.3.4 2.3.5 2.3.6 2.4

Science, scientific literacy and attitudes to animal research 2.4.1 2.4.2

2.5

Gender Early experience with animals including pet ownership Age Religious affiliation Lifestyle variables Liberal outlook

Surveys What scientists think 2.4.2.1 Education will reduce hostility 2.4.2.2 Animal welfare is important 2.4.2.3 Summary

Science literacy models and definitions

7 8 9 9 10 10 11 12 14 15 16 17 17

v

2.6

Chapter summary

19

Chapter Three: Research methodology 3.1

Introduction

20

3.2

Research method and design

22

3.3

The sample

24

3.4

Limitations of the research methodology

26

Chapter Four: Analysis 4.1

Introduction

28

4.2

New Scientist database content analysis

28

4.2.1 4.2.2 4.2.3 4.2.4 4.3

Discourse analysis of attitude surveys 4.3.1 4.3.2 4.3.3 4.3.4

4.3.5 4.3.6 4.3.7 4.4

Braithwaite and Braithwaite survey: example of ‘dissociation’ through different treatments of types of animal use Driscoll survey: experimental animals are given primacy thereby promoting identification with them Herzog and Bowd surveys: different construction of animals in science and agriculture, different construction of contexts Plous, Takooshian and Furnham and Pinder surveys: construction of the ‘suffering animal’ in the science context; and negative construction of science Pifer surveys: misinformation on types and use of animals in science Nibert survey: confusion on ‘animal rights’ New Scientist MORI (1999) survey: ambiguities and variability

Features in common of the attitude surveys 4.4.1 4.4.2 4.4.3 4.4.4 4.4.5 4.4.6

4.5

Table 1: search sets of industry terms Table 2: search sets of animal species Hierarchy of concern in social science literature How one animal is shifting status: the pig

Absence confers meaning Lack of definition for ‘animal rights’ Negative weighting of science, difficulty of describing science and use of the deficit model Suppression of variability in responses through survey construction Dissociation and the ‘social matrix’ Summary of survey instrument analysis

30 32 33 33 34 34 37 39

42 46 47 47 48 48 48 49 50 50 51

How secondary sources characterise animal experimentation 51 vi

4.5.1 4.5.2

Setting up the framework of debate Discursive resources of images: pictures are worth a thousand Words 4.5.3 Questioning priorities: effects of omission 4.5.4 False dichotomies 4.5.4.1 Scientists as sociopaths 4.5.4.2 Creating ‘straw men’ and omitting wider ethical concerns 4.6

Inconsistency and contrary values 4.6.1

Rhetorical effects of invoking inconsistency 4.6.1.1 Hills survey: seeing below the surface 4.6.1.2 Braithwaite and Braithwaite survey: deliberate exploration of inconsistency 4.6.2 Summary of inconsistency analysis 4.7

Chapter conclusion

51 52 53 53 53 54 54 55 55 56 57 57

Chapter Five: Discussion 5.1

Introduction

59

5.2

Hierarchy of concern and the ‘social matrix’

60

5.2.1 5.2.2 5.2.3

Status conferred by context Type of use of animals The status of the discipline of science

60 61 61

5.3

Moral status of animals as indicated by use of the term ‘rights’

61

5.4

The depiction of science and scientists: ‘Science is cruel’

62

5.5

Pragmatism is not incompatible with concern for animals

62

5.6

Implications of findings in this study

63

5.6.1 5.6.2 5.7

Misinformation and stigma Informed public awareness

Recommendations for future study 5.7.1 5.7.2

Interviewing the general public 64 Interviewing social scientists

63 63 64

65

5.8

Limitations of study

66

5.9

Conclusion

66

References

67

Appendices

vii

Appendix Ia:

New Scientist database search data Table One: comparison of science and agriculture with respect to ‘animal rights’ and ‘animal welfare’

71

New Scientist database search data Table Two: comparison of different animal species with respect to ‘animal rights’ and ‘animal welfare’

72

Appendix II:

Braithwaite and Braithwaite survey instrument as published

73

Appendix IIIa:

Driscoll survey instrument as published

75

Appendix IIIb:

Driscoll survey instrument arranged by means for items

76

Appendix IV:

Herzog survey instrument

77

Appendix V:

Bowd survey instrument

79

Appendix VI:

Plous survey instrument

81

Appendix VII:

Takooshian survey instrument

84

Appendix Ib:

Appendix VIII: Furnham and Pinder survey instrument as published Appendix IX:

Day’s book title page

85 86

viii

Chapter One: Introduction 1.1

Background to the Study

The use of animals in research has been a controversial issue, particularly in the last 25 years. Scientists and university public relations people tend not to speak of the use of animals in research (Rowan, Loew, & Weer, 1995). The result is that the general public know little or nothing about the role of animals in science and medicine. Animal rights advocates have therefore been in a strong position to fill the void with their ideas.

Animals are used in research to develop and test ideas in physiology; study pathologies; develop drugs; and ensure that medical and cosmetic products are safe for human and animal use. Two thirds of the Nobel prizes in medicine in the 20th century were awarded for research using animals (NABR 2001). However, over the last twenty five years animal rights advocates have actively and successfully campaigned to reduce animal use in science and to increase animal welfare regulations. They have used legal and illegal means to do this, ranging from lobbying politicians and letter-writing campaigns to scientific institutions, to death threats to scientists, laboratory break-ins, theft and release of animals and arson (Breo, 1990; Day, 2000; Rowan et al., 1995). The logical extreme of the animal rights advocates’ philosophical position requires that no animals be used for scientific (or indeed any human) purpose. If this came about, scientists and others argue, it would slow down or prevent the acquisition of certain types of knowledge, and the development of applications deriving from that basic knowledge. The logical extreme of the scientists’ position is that use of animals for research is useful and ultimately necessary for increased human and animal welfare (Singer, 1993).

Over the last twenty years, social scientists have studied attitudes to animals generally and specifically to animals in research of the following groups of people: animal rights activists, animal rights supporters, scientists who use animals in their research, students and the general public. They have found that the public support for the use of animals in research is much less than it was fifty years ago (Rowan et al., 1995), and have tried to discover why.

1

It seems from many surveys that several variables influence people’s attitudes to animals. The surveys’ results suggest that one is more likely to be against the use of animals in scientific research if one is female, young, vegetarian, has owned or does own pets and/or does not take prescription drugs. Attitude to science and level of knowledge of science do not consistently influence or vary with attitude to the use of animals in research. None of these demographic variables have changed a great deal over the last 25 years. The question arises, what does account for the apparent decreased support of the scientific use of animals in research?

1.2

Statement of the problem

Means determine ends. The instrument used to measure attitudes can to some extent determine the outcome. Is the measured decrease in support of the scientific use of animals to some extent an artefact of the survey instruments?

For example, attitude researchers have seldom asked the public what they understand by the phrase ‘animal rights’. Therefore it is not possible to know from the surveys whether the public supports increased attention to animal welfare, or outright abolition of animal use. The latter is unlikely since approximately 95% of those surveyed eat animals. The only clear idea to emerge (as it were incidentally) from the studies is that the public’s concept of the ‘rights’ of animals does not protect animals from being killed for food.

Some studies have identified inconsistencies in attitudes amongst animal rights activists and the public to the use of animals in research and other areas, but either have not developed or commented on them or interpreted them in terms of imperfectly understood motivations. Many studies embody inconsistencies in their survey instruments, with different treatment of items relating to science, agriculture and other uses of animals. This can skew results.

This study analyses discursively the content of the survey instruments used to measure attitudes to animals. The descriptive words used, reference to type of animals and to specific uses can indicate tacit priorities and biases in the instruments of which the researchers themselves may not be aware.

1.3

Purpose of the study and research questions 2

The purpose of this study is a meta-analysis of the work done on surveying attitudes to animals in research. The research questions addressed are whether there are inconsistencies and biases present in the research on attitudes, and if so, what they are. The study discursively analyses the papers and survey instruments of people doing research on attitudes to animals in general and in scientific research. It also includes books written as popular digests of the subject. To approach the question from a different angle, the study includes a broad content analysis of one popular science magazine, New Scientist.

Research questions and implications:

1.3.1

Do the survey instruments of attitude researchers have a bias, whatever it might be? If so, is it because attitude researchers’ priorities and assumptions about the public’s understanding of science and scientific use of animals have influenced their design of questionnaires and surveys measuring attitudes to animals in research? The crucial problem with bias is that it will influence results and in part manufacture the attitudes the instruments set out to measure.

1.3.2

Are respondents asked to define what they understand by the phrase ‘animal rights’? If not, the conclusions drawn from survey data are weakened as it is not possible to know whether people support improved animal welfare, or outright abolition of animal use. Without knowing what people understand by animals having rights, the debate cannot progress.

1.3.3

Have the priorities of the animal rights movement influenced the attitude researchers or more fundamentally, in discourse analytic terms, are there features of an underlying “social matrix” that are possibly generating any evident biases?

1.4

Significance of the study

There is apparent decreased public support for the use of animals in research though a majority still believe that animals are useful in scientific research and its applications. Surveys show that about 75% of the US public accept the use of animals in research while about 65% actually support the practice. This is down from 85% support in 1949 (Rowan et al., 1995). One Australian survey showed a support rate of 69% (Hills, 1994). The illegal tactics of har3

assment, threats and intimidation of scientists by animal rights activists had dropped off by the mid 1990s (Rowan et al., 1995) but have not ceased (Lancet, 2001). If the animal rights movement’s focus and views on animals in scientific research are accepted, whether intentionally or not, as the main issues on which to base surveys, the data collected from the surveys will fail to give due attention either to the beneficial outcomes of scientific research using animals or to the welfare of the vast majority of animals used by humans. Furthermore, the unintended consequence of publication of results from biased attitude surveys may be to promote misinformation about science, its use of animals and what people really do think on the issue, with real consequences. For example, a widespread stigma remains attached to using animals in scientific research. The Australian airline Qantas had until October 2001 a ban on carrying animals intended for research use (Vidyasagar, 2001).

1.5

Limitations of the Study

1.5.1

Sample

The analysis is restricted to Anglophone, textual discourse including research articles on attitudes, books and an electronic archive of a popular science magazine. The principal objects of analysis are survey instruments. Not all the papers publish the data collection instruments, however, so there is an effective assumption that those survey instruments available for analysis are representative of the corpus. It may be argued that I have used too small a sample (11) of survey instruments in this study. However, small samples are usually not a problem when using the technique of discourse analysis. Potter and Wetherell (1987) argue that small samples are “generally quite adequate for investigating an interesting and practically important range of phenomena” (p.161) as they assert it is possible for a few people to generate a large number of linguistic patterns. The limitation of the discursive material of the surveys is that they are formal and highly structured. They are less rich, detailed and variable than talk. This study has not included interviews as discursive material for analysis.

1.5.2

Technique of analysis

4

The technique of discourse analysis concentrates on the text, in and of itself. It does not take into consideration anything extraneous to the text, including hypothesised internal psychological or cognitive structure or mechanism such as an attitude or an intention. Therefore, this analysis is not concerned with what social scientists might mean with respect to their survey constructions, or what their own attitudes could be, only with the effects of the language used. The New Scientist discourse analysis provides data on whether the terms “animal rights” and “animal welfare” are mentioned in the contexts of science and agriculture. In itself, it is a simple counting exercise, indicative only of a trend, which is a limitation. However, it provides independent confirmation of similar tendencies with respect to representations of animals in the attitude survey instruments.

This study does not answer the question of whether there is in fact a genuine decline in public support for the use of animals in research.

1.6

Overview of the Study

A further four chapters comprise this thesis. Chapter Two contains a literature review. Chapter Three outlines the method of the study. Chapter Four has a simple analysis of a scientific magazine database and detailed content analyses of survey instruments and secondary sources. Chapter Five discusses the results, makes recommendations for further work, and concludes the study.

5

Chapter Two: Review of Related Literature 2.1

Introduction

Chapter Two describes the literature on attitudes to animals in research from several perspectives. The primary source material is the surveys conducted by psychologists to study and analyse public attitudes to animals in research. Since the topic includes both animals and science, there is a section describing how psychologists have tried to tease out the relationship of attitudes to science and to animals. Scientists who work with animals also have views on how and why public opinion is formed regarding the subject and how to remedy the problem of public misunderstanding through an increase of scientific literacy. I have provided a summary of scientists’ opinions. The concept of scientific literacy is not simple, so there is a section detailing the different models of scientific literacy, and how they inform both survey construction and wider discussion of the issue of use of animals for scientific research.

2.2

How attitudes to animals in research came to be a subject of investigation by social scientists

The 1980s saw a massive rise in animal rights activism, particularly with respect to the treatment of animals in scientific and medical research. There were death threats to scientists and hundreds of laboratory break-ins doing up to $10,000,000 worth of damage (Day, 2000). Membership of animal advocacy organizations swelled into the hundreds of thousands, and their budgets to millions of dollars, enabling successful lobbying of governments for changes to animal welfare legislation with regard to scientific research (NABR, 2001; Rowan et al., 1995). Scientific research using animals was, and continues to be, under threat (Lancet, 2001). As a consequence of this social phenomenon, social scientists began studying attitudes to animals in research amongst animal rights activists, scientists and the general public. Therefore, there is a considerable body of literature on the public’s attitudes to animal use in general and in particular to scientific use, generated mostly by US researchers, from about 1980 onwards.

2.3

What the surveys have sought to find out 6

The attitude surveys have tried to discover variables influencing attitudes to animals, including gender (Aldous, Coghlan, & Copley, 1999; Broida, Tingley, Kimball, & Miele, 1993; Eldridge & Gluck, 1996; Furnham & Pinder, 1990; Herzog, 1996; Mathews & Herzog, 1997; Nibert, 1994; NSB, 2000; Pifer, Shimuzu, & Pifer, 1994; Pifer, 1994, 1996; Plous, 1996b), age (Nibert, 1994; NSB, 2000), pet ownership and early experience with animals (Bowd, 1984b; Braithwaite & Braithwaite, 1982; Driscoll, 1992; Jamison & Lunch, 1992; Paul, 1995; Paul & Serpell, 1993), knowledge of animal products (Bowd, 1984b; Plous, 1993), religious affiliation (Bowd & Bowd, 1989; Driscoll, 1992; Kruse, 1999), environmental concern (Kruse, 1999; Pifer et al., 1994), scientific knowledge/literacy (NSB, 2000; Pifer et al., 1994; Pifer, 1996; Plous, 1997; Takooshian, 1988), education level (Furnham & Pinder, 1990; Pifer, 1994), concern for civil liberties (Nibert, 1994), early home influences (Pifer, 1994), and personality factors (Mathews & Herzog, 1997). Lifestyle variables have also been studied, such as vegetarianism, wearing of fur or leather clothes, taking prescription drugs and animal hunting (Aldous et al., 1999; Plous, 1997; Takooshian, 1988).

2.3.1 Gender

Almost all the studies found a difference, usually statistically significant, with respect to gender. Men were more supportive than women of the research and teaching that used animals, and women were more supportive than men of animal protection measures. Eldridge and Gluck (1996) did a study on gender and animal research, and found: … women college students in this sample [of 139] seemed to be more willing than men to make personal sacrifices to protect animal lives, were more likely than men to question biomedical research with animals on scientific grounds, reacted more emotionally and empathetically to the suffering of animals, and were generally more concerned about the plight of research animals. Women also saw a need for more restrictions on animal research. (p.249) Eldridge and Gluck’s summary covers exactly what other researchers found in varying degrees. For example, Kruse (1999) found that more women display higher levels of animal rights advocacy, support for the extension of moral rights to animals and opposition to the use of animals in medical testing than do men (p.185). Mathews and Herzog (1997) found gender was a significant predictor of attitudes towards animals (p.4). In their study, women had higher scores i.e. more concern for animals, than men and accounted for 19% of the variance in 7

scores. Pifer (1994; 1996) in a longitudinal study of about 3000 adolescents, surveyed at age 12 then again at age 17, found that “the consistent supporters of animal research were predominantly male (72%), while the consistent opponents of animal research were predominantly female (70%)” (p.300). Pifer, Shimuzu and Pifer (1996) assessed public attitudes to the use of animals in research across 15 nations (approximately 17,000 participants), and found that in all of them more women than men were opposed to research using animals. 10 of the 15 nations had statistically significant differences. Aldous et al. (1999) found that of their sample of 2000 British people, a total of 64% disapproved of animal experimentation. The gender breakdown found 71% of women disapproved compared with 57% of men, similar to Pifer’s findings (1994; 1996). An earlier British study, by Furnham and Pinder (1990) found that gender is “a powerful and consistent correlate of attitudes with females being more anti- than pro-animal experimentation” (p.447). Hills (1994) found that 7% more men than women in her survey of 303 Perth residents supported animal research and another survey, of US college psychology majors, showed that 15% more men than women supported animal research (Plous, 1996b). Likewise, Bowd and Bowd (1984) found that women registered more humane attitudes (p.22). On a slightly different theme, Herzog (1996) found that women “were significantly more likely than men to ascribe mental capacities to animals” and that those “who attribute high levels of mental capacity to animals were more concerned about animal welfare issues” (p.19). With respect to membership in animal rights organizations Jamison and Lunch (1992) found that it was preponderantly female: in their survey the breakdown was 68% female, 32% male (p.445).

2.3.2 Early experience with animals including pet ownership

Studies found that early experience with animals including pet ownership influenced attitudes towards animals, though it accounted for very little of the variance in scores. Jamison and Lunch (1992) found that 87% of animal rights activists approved of keeping pets, and that “intensely emotional experiences with pets were a significant mobilizing force in the activists’ lives” (p.448). Driscoll (1992) found that 70% of her sample of the general public owned pets. The pet owners, similar to the Jamison and Lunch sample of activists with respect to their emotional attachment to their animals, rated use of animals in experiments significantly less acceptable than non-pet owners (p